[unreadable] Analytical ultracentrifugation is a rigorous method to characterize the stoichiometry, energetics and regulation of biomolecular interactions. Recent advances in instrument design and analysis methods have expanded the scope of sedimentation equilibrium measurements. However, this method has typically been applied to self-associating systems; studies of interactions between dissimilar partners (hetero-association) have been more limited by the complexity of experimental design, data analysis and the lack of appropriate software. Thus, many of the protein-protein interactions being discovered by high-throughput proteomics methods are not readily accessible for quantitative characterization by analytical ultracentrifugation. We propose to develop an integrated software package and new cell designs to facilitate studies of complex macromolecular interactions by sedimentation equilibrium. The package will simplify processing of primary data, fitting of experimental data to self- and hetero-association models by nonlinear least-squares algorithms and will integrate fitting results and graphical output. Analysis of hetero-associating systems is particularly challenging due to the complexity of the fitting models and contributions from multiple species present in solution. Our package will take advantage of the capabilities of the XL-I centrifuge by implementing global analysis of multiple data sets obtained at different sample concentrations, rotor speeds and detection methods (multiple absorption wavelengths and interference optics). In addition, the software will implement constraints on adjustable parameters as well as conservation of mass/signal algorithms to further confine fitted parameters to physically reasonable values. We also propose novel centrifuge cell designs to redistribute mass away from the cell boundaries so that errors in optical detection at the meniscus and base will be less significant in the conservation of mass calculations. The software will include modules for simulation of experimental data and error analysis routines to define the confidence intervals of the fitted parameters and permit visualization of the error surface. Our goal is to develop a robust, software package with a sophisticated user interface that is suitable for wide distribution to the scientific community. The analysis package, along with documentation and tutorials, will be freely available over the internet. [unreadable] [unreadable]